Hongru Hu
Hongru Hu
Hi, I was wondering for your original publication, how large the input dimension (number of peaks) the SCALE model can accommodate. Sometimes the scATACseq give us over 100k features/peaks, I...
Hi, I was wondering if I would want to build a h5 file as the training input for the BABEL model from a count matrix + a metadata Dataframe, are...
Hi, When I looked at some examples of getting the pretrained models, I saw: ```python parameters, forward_fn, tokenizer, config = get_pretrained_model( model_name="500M_human_ref", embeddings_layers_to_save=(20,), max_positions=32, ) ``` ```python parameters, forward_fn, tokenizer,...
to extract gene embeddings, I was trying to use three models: `model_types = ['whole_human', 'human_brain', 'human_blood']`, which means, I have 3 `vocab.json` files I briefly checked the files themselves by...